LARS: Light Augmented Reality System - Benchmarking Tests Video

The benchmarking video demonstrates the performance of LARS as a closed-loop mixed-reality system. Three evaluations are presented: scalability, where up to 399 virtual robots are rendered and tracked; tracking accuracy, measured with 100 real Kilobots under different processing loads (logging only vs. logging plus video recording); and robustness, tested under varying brightness levels and projected visual noise. Together, these results show that LARS remains responsive, accurate, and stable across diverse experimental conditions.

本基准测试视频展示了LARS作为闭环混合现实系统的性能表现。本次测试共包含三项评估：其一为可扩展性测试，可渲染并追踪多达399个虚拟机器人；其二为跟踪精度测试，使用100台真实基洛博特机器人（Kilobots）在不同处理负载下开展测试，处理负载分为仅日志记录与日志记录加视频录制两种模式；其三为鲁棒性测试，在不同亮度水平以及投射视觉噪声的条件下开展测试。综合上述测试结果可见，LARS在各类实验环境下均能保持良好的响应性、精度与稳定性。